A conference system includes a central unit in connection with several delegate units for the participating delegates. The central unit serves to combine audio channels from various delegate units and to distribute the received audio and data channel and combinations of these channels. In the prior art solutions these delegate units are connected to the processing means of the central unit in a daisy chain way. Each central unit has a number of ports to which branches of delegate units can be connected. Optionally one or more extension units are provided to extend the amount of branches of delegate units in the conference system. FIG. 1 illustrates a typical set-up as known in the art. Every delegate unit in the system has two ports, an OUT port and an IN port, typically implemented with differential lines. The OUT port can be represented as an external transmit pair and an internal receive pair (see FIG. 2A) and the IN port as an external receive pair and an external transmit pair (see FIG. 2B). Note however that external or internal pair only constitutes a possible representation. Every IN port should be connected to an OUT port.
Given this conventional set-up the following problems can occur. During the installation of the system, a wrong connection can be made by connecting two IN ports or two OUT ports with each other. The cabling from left to right or right to left is different or the cables can cross each other.
As delay sensitive information signals such as audio and video signals are to be transported over the network, the interconnected units of the conference system must be synchronized. By achieving a synchronized network the data handling at each sub system is minimized, which e.g. avoids the need to resample the data.
Patent document U.S. Pat. No. 5,930,237 discloses a video conference system with ATM cell reuse. The video conferencing terminals are arranged in a daisy chain with the head of chain terminal connected to the ATM network. Each terminal has a video program input port with a unique ATM address and a plurality of video program output ports, each provisioned with one of the video program input port addresses.
Patent application EP1717978 A1 is concerned with time synchronisation, deterministic data delivery and redundancy for cascaded nodes on full duplex Ethernet networks. The proposed solution allows precisely synchronizing clocks in separate nodes on a communication network by adjusting timestamps and related data in network messages. Daisy chain connection of the nodes can be established. Redundancy is provided by extending the daisy chain to a ring topology. A supervisory device is then provided that has one master clock with two specialized ports and a specialized signalling protocol for obtaining redundancy. During normal operation, the supervisory device monitors traffic by sending special packets. Simultaneously, the supervisory device and end nodes monitor link status of their ports periodically and the end nodes notify the supervisory device in case of failure of a port through other port. When the supervisory device detects or is notified of a network failure, it broadcasts this status to all nodes through two different messages on its two ports. Hence, in this disclosure a failure is detected or notified via a central supervisory device.
U.S. Pat. No. 4,442,502 relates to a digital information switching system in the form of a family of information switching exchanges for effecting the connections from person to computer and from computer to computer within the business office. The proposed solution allows to add and to change in an easy way both telephone instrument features and system features by changing system table entries or the addition of software modules. Automatic detection of device failures maintains the system in operation at all times. Thanks to immediate automatic notification of occurrence of faults, faulty devices or components can be immediately replaced.